A field experiment was conducted in a farmer field in Alneaamia region/Fullujah/Anbar governorate, during the winter season 2012-2013. Seed of four cultivars of durum wheat (Abu-Gariab, Aliraq, Tahady and Alezz) was planting under three planting rates (100, 130 and 160 kg.ha-1). Split plot arrangements in RCBD design were used. The objective of this study was to determine the more effective traits in grain yield, which were considered it as a selection index to improve grain yield in wheat. The result of analysis of variance showed that there is a highly significant difference for all studied traits for all planting rate. The value of genetic variance was highly than environment variance for all traits and it was increase with increase of planting rates exception of leaf area which was decrease of its genetic variance with increase of planting rate. Number of spick.m-2 gave highly broad sense heritability (0.956) for planting rate (130 kg.ha-1) and number of grain.Spike-1 (0.902) for planting rate (160 kg.ha-1).
The value of standard coefficient variance and genetic, phenotypic covariance was contrast with planting rate, there is highly value of its in grain yield and number of grain.spike-1. There is positive genetic, phenotypic and environment correlation concern between yield and number of spick and number of grain and there is a negative correlation between yield and leaf area and weight of 1000 kernel at planting rates (130 and 160 kg.ha-1) while there is a positive and highly correlation of environment between yield and number of grain.spike-1 and weight of 1000 kernel at the three planting rate. The path coefficient analysis showed are the number of spick have a highly direct positive effected (1.339, 1.273 and 0.776) at the three planting rate respectively, as well as the total of effects (0.982, 1.095 and 0.965). The number of grain spick-1 gave highly positive indirect effects (1.365) among number of spick at 130 kg.ha-1